MFR PAPER 1095

Deepwater Shri mp Trapping in the Hawaiian Islands

PAUL STRUHSAKER and DONALD C. AASTE D

ABSTRACT-The results of preliminary deepwater shrimp trapping sur­veys in the Ha waiian Islands consisting of 82 sets involving tota ls of 306 traps and 3,960 h of fishing time in the 75 to 450-fathom ( 135-825 m) depth range are discussed. A variety of traps , bait containers, and baits were tested. Attempts to capture the penaeid shrimp, Penaeus marginatus , in substantial numbers were unsuccessful . More success was experienced in trapping the ca ridean shrimps Heterocarpus ensifer and H. laevigatus . For these species, it was found that traps covered with burlap cloth out­fished uncovered traps by factors of 2.5 to 10. H. ensifer was found to be most abundant in depths of 200-250 fathoms (365-455 m) where catches with optimum trap and bait combinations ranged from 15 to 63 pounds (6.8 to 28.6 kg) per trap for sets of less than 24 h duration. Seventy traps set overnIght during spring 1973 off the south coast of Oahu produced a total of 1,015 pounds (460 kg) for an average of 14.5 pounds (6.6 kg) per trap . The larger H. laevigatus was found to be most abundant in depths of 240-375 fathoms (440-685 mY, where catches ranged up to 34 .8 pounds (15.8 kg) for 4 traps. It is speculated that the various species of Heterocar­pus represent an unexploited world resource of considerable magnitude .

INTRODUCTION

During 1967 and 1968 the Nationa l Marine FIsheries Service (NMFS) de­voted four cruises of the RY Town­

send Crolll\\ ell to demersal resource surveys in the Hawaiian Islands . The primary ampling gear uti li zed were shrimp trawls. a lthough a few shrimp trap stations were effected. These sur­veys demonstrated that the penaeid shrimp. PenllellS IIllIrginallis Randall. was available in modest amounts and

amenable to harvest by shrimp trawls in dept hs of 60-125 fathoms ( 110-230 m)1 (Yoshida . 1972: Struhsake r . Yoshida. and Shomura2 ). Two species of caridean shrimps. H elerocllrp1l1

el/liji-r Milne-Edwards and H . lael'igll­

IllS Bate (Fig . I), were a lso taken in

.~ Metric eq uivalents are approxim ate. - Struhsaker. P., H.O. YoshId a. and R.S. Shomura. Exploratory shrimp trawling in th e H awaiian Islands. unpubl. ms .. Southwest Fisheri es Center. Honolulu Laboratory. Honolulu . HI 96812.

Paul Struhsa ker and Dona ld C. Aasted are on the sta ff of the South west Fisheries Cen­te r , National Marin e Fisheries Service , NOAA, Honolulu , HI 96812.

moderate numbers with th e trawls between 150 and 400 fathoms (275 and 730 m) . However. the initia l in ­vestigations indicated that trappi ng. rather than traw ling. was perhaps a more effecti ve method for capturing th ese latter species.

larke (1972) conducted trappIng urveys during 1969 and 1970 at sev­

era l locales off the Hawaiian island of Oahu . The sampling gear consi ted of evera l trap type . but all were un­

covered traps constructed of 0.5-inch ( 13-mm ) mesh wire sc reens . This wo rk provided data on the biology and depth distributIon of H . en.liji-r a nd H . l{/ e\'i~(/llI\ and indicated th a t H .

ensiler mi ght be present in sufficient quantities to support a commercial fishery.

Further trapping trial were inter­mittently conducted by M FS during 197 1-73 at various localities in the H awaiian Islands . These trials were of a preliminary nature . and various combinations of trap. bait containers. a nd baits were utilized . Most of the data were obtained during fall 1972 and it became apparent at this time that burlap-covered traps were more effective in capturing H . ensiler than were uncovered traps . Attempts to trap P. lIlarginllllls in sub tantia l num­ber were u nsuccessful. Fol low-up cruises wit h more extensive trapping

FIgure 1.-A tr8lh specimen ot Heterocarpus ensi'er (A, lelt) about 115 mm total length . Note ocellus on the sIde ot the cephatothorax ("head " ) and the strong backwar dly projecting spines on the th ird and fourth segments of the abdomen ( " tall " ) . A tresh specimen ot Heterocarpus laevlgatus (right , B) about 175 mm total length .

24

Figure 2.- A covered square trap . Th is trap type produced the best catches 01 H ele,oca,pus ensifer .

tri a ls sch eduled fo r 1973 we re ca n­ce ll ed .

GEAR AND SAMPLING EFFORT

Gear

T he "squ a re tra p" (in reference to its squa re cros -secti o n) produ ced th e bes t ca tches o f H eleroca r p li s. It mea­sures 2X2X 4 fee t (60x60x 120 cm) (Fi g. 21. The fra me is const ructed o f ~8-in ch ( IO- mm ) di a meter stee l ba r a nd i encl osed with 0.5-in ch squ a re mes h ga lva ni zed wire screen . The sides. top . a nd botto m are th en cove red with burl ap . Th e tunne ls (one at each end ) are a lso co nstructed with the screen (but are uncove red ) and taper to approx ima tely 3- in ch (75 -mm ) di a meter openings about o ne-third th e di sta nce in fro m each end Cove red and unco vered t raps of si mil a r con­stru ction . but enclosed wi th I-i nc h (25 -mm ) (bar-to-ba r) he:\.agona l wire creeni ng were a lso tes ted .

" F lat traps" (F ig. 3). can tructed ~imil a r l y to the qua re traps. but mea-uring I X 2 X 4 fee t (30X60X 120

cm ). we re utili zed Ill OSt oft en in the shal lower depth sun eyed . It was fe lt th at the lowe r ramp angles of the tun ­nel m igh t pro \ e to be mo re effeclI \ e in capturing P. IIwrgill<lllll. apti\ e specime n~ of P. IIl<1r~ill<lf1l1 held In large tan I.. as ho re tend d to ma inta in c1o ' e cont act \\ ith the botto m a nd u ' ualJ\ did not undergo e \.pl o rato l1 \\ immtng heha\ lor off the botto m

such a' h a~ heen ob e n ed fo r las l..an

pandalld "hnmr" In the \ I~IOII\ 01 trar\ ("- ..:""I..:r. Illhlll

( o ll ap\lb l..: n.lund \hnmp Ir..lr' (/lg

4) were aha te,ted to a Ilmll..:d dcgru.: Th e~ are 14 Inch..:, (3(-, cml high . 24 In che (60 cml In diameter al th..: I\lp. 34 In ch c~ (X6 cm) In diam..:t..:r at the bott o m and ha,e \hallcl\.\ tunn..:l" l..:..Id­in g to four entrance" The \I...:h­bi ng con~i.,t, of I 25 -I nch (~I -mm l

stre tc hed mc.,h n~ Ion . Fina ll y . an expenmen tal collap"l­

bk "tent t rap" (Figs. 5. 6. and 7) wa, teskJ o n one occas ion. Each 01 the three frame ma l.. ing up the hottom. s ide. a nd li d of the trap mea~ure,

2X4 feet (60x 120 cm) and I" en­cia ed with 0.5-i nch wire mesh "creen . Th e three frame are joined with ~pllt

links of a s ize th at perm it folding of

th e frames and are covered with a si ngle p iece of bu rlap. A bridle attach­ed to the corners of the lid runs through the outside corners of the botto m frame (Fig . 6) . When clo\ed. th e li d is secu red with a rubber strap a nd S-hooks. The triangu lar end frames a re attached to the bottom of the trap by sp lit links a nd have two ha lf- lin ks welded on the leg articu lat­ing wit h the side frame (Fig 7) . The tunne ls are const ructed With 1.5-lnch (38-mm) mesh ( tretched) n)- Ion web­bi ng and are secured to entrance n ngs 3 inches (75 mm) in diameter T he tunne ls are held in place by a rubber strap and S-hook (Fig . 6) .

Several bai t containers were teo ted Containers constructed of O.5-lnch mesh screen meas uri ng about 4 X 4 X 6 inc hes (lOX lOX 15 cm) were gen­erall) su perior to plastiC containers of the ame size perforated With 025-inch (6-mm) ho les or bait wrapped In cheese clot h . F rom I to 2 pounds (0.5-1 kg) of ba it were ~ufticlent for

Flgur. 3.-A" unc:o....red a'trap

15

F,gure 4 - A coil pl lb le I hrlmp pOI

"..:1\ ,I f 12-1 4 h Th were ,u'r"n ded n"ar t h~ trar

E-ach trap ba I

bndle of () 2"-II1c.h

of the trap 1,lndl n

,et It \\a, al 0 th

out throu gh the t:

Ing bad, 1 he rrlJ center f,lr dtt dc.h Itne fh c ground

on the order of 50-150 fathoms (90-275 m) greater than the depth of water being fis hed. A cha in weight is snapped onto the buoy line abou t 25-50 fathoms (45-90 m) below the end of the last section to a sure that the polypropylene line does not float at the ea surface. The set I retneved with a line-hauling power blocI-..: th e 100-fathom seCllons are stowed di­rectly into their containers as the line comes aboard.

Because of the buoyancy of the polypropylene line, minimal flotation for the marl-..er buoy IS required. Our buoys consisted of either simple bam­boo pole with two or three pia tiC

Figure S.-A collapalble tent trap ready 'or 'Iahlng.

longl i ne float s or a more e laborate buoy consisting of an I H- foot (5-m) fl ag and light stafl o f 2.5- lnch (64-mm) di amet er condUit piping which fits into a weighted frame which 1\ fl oa ted with a large Inn er tube. The~e buoy~ rema in ed at the surface even In areas where the currents approached 3 I-..n ots. A 4-fathom (7-m) pi I-..up line with a small float IS a tt ached to the buoy .

Distribution of Sampling Effort

£-rom March 1<:171 to Augu\t 1971, 82 exploratory shnmp trapping \et~

InvolVing totals of 306 traps and about 3,960 h of fishing lime 'Were elfected In "anou areas of the Ha'Walian h­lands . The dlstnbutlOn of sampling effort by depth 'Was as follo'Ws 75-11() fathoms (135-23'i m). 24 ets, 130-200 fathoms (23'i-365 m), 2 sets, 200-300 fathoms (365-'i50 m), 47 ets, 100-400 fathoms (550-710 m). 7 sct~. 400-4'i0 fathoms (730- 25 m), 2 \ets Most ampllllg effort 'W e\pended III depths

of 75-130 fathoms and 200-300 fa­thoms 'Where P. IlI1Irs.:illllllll and H I! III I./er , r~pecllvel), are mo t abun­dant

Most of the trappl ng effort f r P l11a(s.:lllllIlII tool-.. place III the Pail 10 Channel where thl pecle I most abundant ( truh al-..er et ai, unpubl m .). The deeper depth range ex­plored with the R TOlIlHl!lld Crol11-

II I! II were off th e north coa t of Oahu. the north edge of Pengulll Bank , ff

northern MaUl , and th e west coas t o f the 1\land of Hawaii off Kawa ih ae. Dunng spnng 197\ , 21 set~ were made oil the south coas t o f Oahu With our gear by personnel of the Hawaii Inst itute of Manne BIOl ogy (H fMB J utlllLlng the I V Vallalll M£IId Dunng

ugust 1971, 9 set'i we re made WI th the R V Dal It/ ICIrr l ()rt/()1I In th e l eeward Isl ands of the Hawaiian

rch lpelago nl.!ar ecker hland, fr ench f- ngate Shoal , l. aysan Isl and, and Plonl.!l.!r BanI-..

Because the trappl ng I.!xperi ments werl.! IIltersper<,ed With other ves<,el actiVit ies, most \eh 'Were made at du I-.. and retneved at da'Wn Ho'Wever. 'We do havc a fe'W data for daytime sets We trapped n a vanet) of bot­tom t)pe\ , aVOiding onl) steepl) slop­Ing areas and rough-bottom grounds 'Where thc traps might have been \nagged

RESULTS

Trapping for Penaeus marginatus

ttempt to trap P. IIll1rs.:llIlIIlIl III

sub tanllal number 'Were unsucces -ful. but the sampling effort .... as of a very prellmlnar) nature fter gen­erally negall\.e result obtained dunng spnng 197 I. e\pe n ment 'Wllh capll\. e hnmp 'Were conducted at the IF

Ke'Walo Ba In faci llt ] R pon e of the hn mp 'Wa noted dunng hours of li ght and darkne to baited a nd un ­bai ted flat trap (a ll uncovered) eq ulp-

Figure 6.-A tent trap with ralud lid.

26

Figure 7.-A collapsed tent trap . The end 'ramea 'old outward , over the bottom panel.

ped with a variet y of entrances. Rath e r surprisingly, there were no majo r differences in the catch rates be tween the various combinations of th e trap types and baits tested. These result s were proba bl y confused by a bnorma l temperature and light conditions. A l­though adult P. lI1arg inalll s appear to be negativel y phototactic (Struh­saker et a l" unpubl. ms.), best capture results were obtained by suspendin g a li ght over an unbaited tra p . This resulted in all of the available shrimp (18-20) in a test ta nk 24 feet (7 01)

in diameter entering the trap within 15-20 minutes .

Further field trial s conducted dur­ing fa ll 1972 were inco nclu ive. as ide from determining that flat traps out­fished square traps on th e a me string . Traps containing a deep-sea di ving light captured shrimp on three occa­sions when th e other traps o n th e string did not ta ke any shrimp (maxi­mum s ingle trap catch, five shrimp). Howeve r, the best catch of nin e shrimp for a n overnight e t was in a n uncove red flat trap baited with fi ne ly ground fis h . Three sets of traps fished during dayli ght hours produced low catches simil a r to night sets.

Clarke ( 1972) reported a best ca tch of thi s speci es as 26 individua ls in eight traps in 60 fathoms ( 110 01 )

off th e no rth coast of Oahu . We feel that concerted trapping tri a ls for thi s species might result in a viable fishing strategy that could produce catches of commercia l va lu e for sma ll vesse ls.

Trapping for Heterocarpus ensifer

During fall 1972 seve ra l types of traps , bait containers, a nd baits were tested in various combinations o n set of 4-6 traps per st ring . These pre­liminary trial s were not designed for the results to be subjected to stati sti ca l evaluat ion , but the main trends of th e effectiveness of th e various trap com­binations in capturing H . ensiler were readily apparent .

Square traps with 0.5-inch me h wire screen were more effective th an the other types tested . The catches of square traps with I-inch mesh wire screen were less, apparentl y owing to escapement. Catches from the flat traps and round traps were consistent­ly lower than for th e square traps.

Figure 8 .-A 123-pound (56-kg) catch 01 Helero­carpus ensiler produced by lour traps .

The performance of burlap-covered and uncove red sq uare traps wi th 0 .5-inch mes h were directly compared on eight occas io ns when strings of a lte r­na tin g trap types were set. The cove red traps a lways o ut fis hed th e uncove red traps by facto rs of 2 .5 to 10.

Because mos t catches of H . ellsiler

were made a t ni ght in deep water . it would seem that they are not differ­entially entering th e covered traps to seek she lter. We agree with Bu tler ( 1963) that possibly the covered t raps are rendered more effecti ve in that the ba it scent i concentra ted at the trap entrances. rather th an diffused through th e sides in add itio n to the entrances in th e case of uncovered traps .

Five ge nera l types of bait were tested: fine ly ground fi sh , squid , shrimp (H . ellsiler) , and coarse ly shop­ped fish an d shrimp . F ish a nd shrimp were superio r to squid a nd the coarse­ly chopped fish and shrimp outfished th e fi ne ly ground containers of th ese species . This finely ground bait prob­ably dispersed too rapidly. Bloody and oi ly fis hes such as skipjack tuna (KalslllVo nlis pe/all1is) , boarfish (A 11 -

ligonia spp .) and A 1I/0plIS we re found to be superior to lizardfishes (Sy nodontidae), flatfishes (Both idaeL and smelt (Os meridae) .

Ca tch rates of H . ensiler obtained during fall 1972 were h igh ly variable due to th e variety of traps, containers. and baits tested and the depth sam­pled . At this time the best catches

27

ranged betw een 15 and 63 pound~

(6 .8 and 28 .6 kg) per trap on o\cr­night sets. On e of the better cat(hc~ made a t thi s t ime was off the north hore of Oahu in 240 fathoms (440 OJ)

where 122 .7 pounds (55.8 kg) of If ellSlfer were take n with four square traps baited wi th fish (Fig. 8, Tv,o cove red traps accounted f( r O.h pounds (49 .3 kg). while tv.o unco\ ered traps captured 14.4 pnurds ( 'i

kg) . The best ingle catch of H. (I, f I

was 63 pounds (286 kg, Ir()rr d t h bait ed. cove red squarc trap d nn I

24-h set in 225 fathoms {410 01 nl rth of Maui . Another covered squ",rc trup on thi s trin g. baited \\ Ith ~ I produced 24 pounds (IO.q g. the rema in ing two cO\ereL q traps a lso baited With fish a d ~h produced only 1.5 and lOr (0.7 and 0.5 kg). respe<..tl\cl T great d is pa rity in the catches Inul\.. j

to us th at the latter two traps (\\hl"n were positioned In the middle of thc string) landed on end or v.ere rl"

fi shing for some other reason D uring spnng 1973. two o\ernlght

sets of SIX covered square traps on a string were made In 225 fathoms (410 01) off the north coast of Oahu to pro\ide shrimp for test marketing. One string produced a total of 134 pounds (605 kg) of H. em!}£'/'. v.lth three traps baited with skipjack tuna accountIng for 10225 pounds (46 .'\ kg), and three traps ba ited \\ith v.hite­bait smelt (AlloslI1cru.\ elon~(/llI\) tak­ing 31.75 pou nd (1 4 .0 kg). The sec-

ond string produced a total of I 18.5 pounds (53.9 kg), the three skipjack tuna baited traps producing 80.5 pounds (36.6 kg), while the three whitebait smelt traps captured 38 pounds (17.3 kg) . The tent trap was also tested on this string (baited with skipjack tuna) and produced an addi­tional 12 .5 pounds (5.7 kg) . This was the single te t of thi trap type. and inasmuch as a square trap is about 2.1 times the volume of the tent trap. we feel that this design might even­tually prove to be practical.

During August 1973. six sets of covered square trap baited with tuna were made in the 200-250 fa thom (365-460 01) depth range in the Lee­ward Islands . The results indicated that the peak abundance of H . ellSifer may occur deeper there than in the main group of the Hawaiian Islands . Four traps produced totals of 46.5-50 pounds (21.1-22.7 kg) at each of three sets between 225 and 250 fathoms (410 and 460 01) near ecker Island . A 6-h daytime set here in 200 fathoms (365 01) with two traps produced only 10 indl\·ldual shrImp. Only 2-3 pounds (0 9-1.4 kg) of H. ell.lifer were taken in two overnight sets of four traps in 205 and 215 fatho ms (375 and 390 01)

near Laysan Island. Our covered square traps were

tested in a more uniform and ex tensive manner dUrIng May 1973 by HIMB personnel when 19 overnight sets (of 21) were made in the 200-220 fathom (365-415 01 ) depth range off the south coast of Oahu between Waikiki Beach and Sand Island . Four traps were fished to a string and were baited wit h a mixture of two species of smelt (primarily Thale/chthys pacificlls and, to a lesser extent. Hyp a m esils pretia­.IllS). Catches varied from 2 to 34.9 pounds (0.9 to 15 .9 kg) per trap . Total catch from 70 traps (complete data not avai lable for six traps) was about 1,0 15 pounds (461 kg) for an average of 14 .5 pounds (6.6 kg) per trap . These data also reflect our i 01-

pression that considerable variation occurs between localities and depths and even between identical trap ar­rangements on the same stri ng . This is probably due to local varia ti on in the abundance of H. ensifer and varIatIon in placement of the bait container and in the trap tunnel en-

20,--,---,---,---,--,---,---,---,--.---,---,--,---,---,---,--,---,

Ie

16

I.

12

10

e

6

He1erocorpus enslfer (NONOYIGEROUS N ,156 )

2 I 0 • .~ ~','

"

.,j , • 0°':'-

", , " " 8

, :'1 1' ' :

"

"

~ 55 60 65 70 75 eo e5 90 95 100 105 110 115 120 125 130 135

TOTAL BODY LENGTH (mm I

Figure 9.-Length-weight curve for 156 non­ovigerous Heferocarpus enslfer.

trances occurrIng dUrIng the original constructIOn or after modification due to damage.

These catches of H . em/fer with covered trap generall, exceeded those experIenced by Clarke (1972) with uncovered traps outside Kaneohe Ba, on the northeast coast of Oahu during

ovember 1970. Catches there for nine quantitative ets in the 150- to 250-fathom depth range varied from 0.6 to 11 .9 pounds (0 .3 to 5.4 kg) per trap and averaged 4 .5 pounds (2.0 kg). At that time the be t catch obtained wa a respectable 96 pound (43.6 kg) from eight traps fi hed in 200 fathoms (365 01).

We have taken H. ensifer in depths ranging from 76 to 360 fathoms (137 to 660 01). Clarke (1972) took the species between 80 and 400 fathoms (146 and 730 01) but concluded that it is most abundant between 150 and 250 fathoms (275 and 455 01). Our results indicate a peak abundance be­tween 200 and 240 fathoms (365 and 440 01), but we have few stations in depths between 150 and 200 fathoms (275 and 365 01). Quite probably the depths of peak abundance for H . ensi­f er vary locally accordi ng to varia­tions in the temperature structure. In the main group of the Hawaiian Islands, bottom temperatures are about 14°C at 150 fa thoms (275 01) , 9.5 °C at 200 fathoms (365 01 ), and 8°C at 250 fa thoms (455 01 ) .3

3 Unpublished data, Southwes t Fisheries Center, National Marine Fisheries Service, Honolulu, HI 96812.

28

~ 0-r (!)

W ~

>-0 0 II)

..J

~ i?

20,---,---,---,-------.------,

Ie

16

I. 12

10

6

?oo

Heterocarpus enslfer (OVIGEROUS N' ge)

, ,

. " " I °'0-0. 0.:

D, .o'! • °i

105 110 115 120 125 130 135

TOTAL BODY LENGTH (mm I

Figure 10.-Length-weight curve for 98 ovigeroul Heterocarpus ensi fer.

Clarke (1972) al 0 demonstrated that the 150- to 250-fathom depth range was the zone where the largest size-class of shrimp occurred. The catches in greater and lesser depths were comprised of a significantl y greater portion of maller indi vid ual s. We did not collect extensive length­freque ncy information. but estima­tions of the size composition of the catches also indicate that smaller shrimp are fo und above and below the depths of peak abundance.

Clarke (1972) pointed out that H. ensifer seems to exhibit protandrous hermaphroditism (sex change from male to female with growth), but that the situation seems to be more com­plex than for some of the commercial pandalid shrimps in that a large por-

tion of the males reach near-maximum size without transforming and that many small individuals (below egg­bearing size) are females. A large proportion of the larger females taken are ovigerous (gravid , or "berried" ) throughout the year . Clarke (1972) presented data that indicated that this condit ion varied wit h depth and that the proportions may vary seasonally.

The largest s ize-group of H. ensifer taken in our traps usually weighed about 35 to 45 individuals per pound (75 to 100/kg) with the heads on , although the largest indi vi duals taken run 28-30 per pound (62-66/kg). We obtained total length (tip of rostrum to tip of telson) and weight data for 156 nonovigerous H. ensifer (Fig. 9) and 98 ovigerous females (Fig . 10). Data were obtained from specimens frozen less than a month, thawed , and damp-dried . The fitted exponen­tial length-weigh t curve for the non­ovigerous individuals is: 10glQ W = 0.0162 L - 0.8789, where W = total weight in grams , and L = total length in millimeters ; 1'2 = 0.972. The equa­tion for the ovigerous individuals is: 10glO W = 0.0105 L - 0.1947, where r2 = 0.589 , W = total weight in grams, and L = total length in milli­meters.

We also determined the percentage

20

10

!] i ~1::':" ~ 1

I ~r~"' I I ,: •• ~ : 1

~r lL O,"·~' 1 ~ ~ ~ c ~ ~ ~ 00

PERCENT

Flgu,e 11 .-Helerocarpus ensifer. Weight 01 tail with .hell a. pe,cent 01 total weight : A. Ovlge,­ou. lemale. ; B. Nonovlge,ou. Individual • . Weight 01 tall without .hell al pe,cent 01 total weight : C. Ovlge,ou. lemale. ; D. Nonovlge,ou. individ­ual • .

of the total weight contributed by th e tails of ovigerous and no no vl gerous individuals. bo th with and with out shell and eggs (Fig . II) . The maj o rit y of ovigerous individuals have tota l tail weights making up a bout 45 -55 percent of the total weight, whe reas the tail weights of most of the non­ovigerous individuals were about 40-50 percent of the total weight (Fi g. IIA and B). These data indicate th at the egg mass is about 5 -I 0 percent of the total body weight. With the shell and eggs removed , the tail muscle of both ovigerous and nonovigerous in­dividuals make up about 25-35 percent of the total weight (Fig. IIC and D) .

Trapping for Heterocarpus laevigatus

H. iaegivafll s (Fig . I) is a large species of H elerocarpLis inhabiting a deeper depth range than H . ensifer. Clarke (1972) collected this species between 200 and 400 fathoms (365 and 730 m), while we have obtained it in depths ranging from 2 35 to 450 fathoms (430 to 825 m). We have rather limited coverage in the deeper portion of this depth range, but several catches indicate that this species can be taken in sufficient quantities to support a commercial fishery. Our initial catches with uncovered traps were not encouraging , and were simi­lar to those experienced by Clarke (1972) where yields were usuall y on the order of 1.5 pounds or less per trap .

Some of the better catches of H . iae l'iga ills we obtained with covered traps are as follows. South coast of Oahu : 18 pounds (8.2 kg), three traps, 246 fathoms (450 m); 10 .6 pounds (4 .8 kg) , two traps, 255 fathoms (465 m) . North coast of Oahu : 11.1 pounds (5.0 kg), two traps during a 4-h day­time set, 285 fathoms (520 m ). Kawai­hae , Hawaii : 7 pounds (3 .2 kg), two traps , 238 fathoms (435 m). Leewa rd Islands : 15 pounds (6 .8 kg). four traps, 320 fathoms (585 m ). Pioneer Bank, 34 .8 pounds (15 .7 kg). four traps. 360 fathoms (660 m ), French Frigate Shoal; 7 pounds (3 .2 kg). four traps, 440 fathoms (805 ml. Pioneer Bank Although the da ta are rather frag­mentary, it woul d appear that thiS species is m ost abundant In depth'

29

of about 240 to 37':.. lath 'm, (44 684 m)

The proportion of large and mall in diV idual, of H . Itltl 1~(III1~ tdk.en 10

the traps abo seems tll \ar: bct\\l:~n

catches. bu t the pattern IS nlll yet clear Large indl\ Iuual, of thiS pecIc att a in lengths of 200-220 mm tntJI length . bu t me,t of the InUI\IUUJI compriSing the largest "ze-gr up ra nge from 150 to 17':.. mm tlltal length and average 8- I I InUI \ Idu Is per pou nd (heads on).

FURTHER CONSI DERATIONS

One problem encountered \\lth H. ensife r is cannibalism of d certain percentage of the trappeu InUI\ IdUdl, W e have no precise data on lhl\ phe­nomenon. but rough estimates indicate that as littl e as I percent to as mu..:h as 15 percent of all the InLlI\ldu I

in a trap may be affecteu Sublec,tl Iy. it seemed that the highest at of cannibalism occurred In s t long dura ti on or In trap 'n .... hleh th bait su pp ly was exhausted I a individuals are most ufte., Jftt. and , if ov igerous. the eggs JftU merets are remond hrst T h viscera , and . finally. the tUI ture are a ttacked ThiS rcs..Jit problem of sorting thE. u parti a ll y devoured Indi'dduul contaminate the c,ate" ( canniba li zed and smJI shr most easily effected in seawater .

We poi nt out that the trap these pre liminary tnals c,;ol\ usage, and certainly do nl t what may become the optimu and construction. Probably ., ta iner u tilizing the princIple, cove red trap \\ ith a long taper d u nel would be suitable. 1he me hilt: of the tunnel should be Inl.:rc ... d t about 0.75 Inc'h ((9 ml'11 to pemllt escapement of m II hnmp P Ibl} a trap \\ IIh ;1P entmpc'e on one end and a flat creen on the other end \\ould permit e "apement d m II indj\ Iduals dunng h uhng of the trap thu redu';lng the am unt of nmg on the fJ hermen' part and mg the premature hant: t shnmp.

The tU11 meat of tht' t\\ of H elero( arplls tend to bre

pt: I

d \\n

rap ldl) '" hen th e hrimp are imprope r­I) chilled aboa rd th e ve el. T he ~hrim p mu t be immediate ly placed In cru hed Ice o r chill ed seawater. Thi~ '" III pro VI de an acceptab le prod­uct fo r up to a bout 3 day afte r captu re W e fo und that cookin g the shrimp nght after capture re ulted in chilled she lf life of 3 to 4 days . O ne tes t of Immed iate freezi ng of raw ~hrimp In a bn ne soluti o n a t a bou t - 20 ° resu lt ed In a n acceptable product A fter freezing. th e sh ri mp "'ere placed In a conve nti o na l freeze r and tested a t Interva ls ove r a 2-week. penod Ver} li ttl e a lt was ta ken up b) the shrimp and the product did not seem to de te n ora te during th e bnef te t peri od .

Care m u t be taken not to overcook these specie. BOi ling time o f 2-4 min are adequate for the frozen H . ('/l\;/er. The tail meat ca n be he ll ed easil) for other methods o f prepa ra­tion by bnefh thaw ing in runnin g '" ater or atter a ~O-sec bl anch .

The onl) present comme rcia l fi h­en for a spec ie o f Hererocarpu .1 (f{. reC'dr Bahamonde) ex i ts pri mari ly off (hill.', and to a lesse r ex tent . Pe ru . In thiS hlghl) productive reg io n o f up\\elling, a tra", 1 fi hery p roduce dhout 10,000 metric tons an nua ll y IHancod: and Hennquez. 1968). It has been e ... timated that the to ta l ava il­..tblc grounds are about 1.600 km2 in extent and that production coul d reach "lwut 2() ,OOO metric tons annua ll y ((,ulland, 197 I) ThiS ind icates a n annu,tl production of about 12.5 metnc tlln /k.m l

In H a\\..tll. It appears th a t th e ab u n­dance llf II. ('/IIi!i'r i~ grea tl y under­stlmdted h\ sampling w ith 4 I-foot hnmp tr .. m!s Our ma"\ lmu m ca tc h

rdte Illr thl\ peLies was 240 k.g/k.m2

althllugh mn,t <.dtc.:hes "'ere les than II) pen:ent III thiS It IS perhaps be t

..tt pr~ ent III not specu late o n th e are.1 eflectlvel) lished b) a Single hnll1p (rdp. hut rather point l)Ut aga in

th .. t Ir.lp' are a much more effiC ient me.ln (II h,ln cs t Inc H. ell\l It r 1 n 11.1\\.111 W~ feel th,1I ..In ..Innu;1 \ leld

I I llr 2 metne l on'/"m~ lo r ' H a­"',111,111 \\,llef d,le n,)t 'eem like an unrC,1 f),lhlc ("'llIll .. lte The t )tal area

f the pllillum depth r.mge fl) r H 1/\1 If 111 Ih Ill .. III 'r,lup ,)1 the

li d'" 111.111 I I.Hld h.1 Ihlt heen pre-

cisely d etermined. bu t pro babl y ex­ceeds 1.000 km 2 .

It wo uld seem tha t H . ensiJer, a n a lmost entire ly cosmopolit an peci es occurr ing fro m H awa ii to the western centra l A tl a nti c. is about as a bundant in o ther pa rt s of its range as in H awaii . In th e western A tl a ntic . expl o ra tory trawl catches by th e M FS we re o f th e same magn itude as we re th ose in H awaiia n waters.4 Limited trapping near G ua m in th e weste rn Pacific 5

have produced ca tch es similar to th o e reported he re .

Using da ta prese nt ed by M oiseev ( 1969). it was ca lcul a ted th a t the 200- to I ,OOO-m depth zo ne o f the wo rld conti nent a l a nd insul a r slo pes in regions whe re species of Herero­carpus may occur has a to ta l a rea of abo ut 5 milli o n km2 . W e assum e th a t th e area of th e 200-600 m d epth ra nge is abou t ha lf thi s: 2.5 milli o n km 2 .

Assu ming th a t 10 pe rcent of thi s a rea is fi shabl e and suppo rts concentra ti o ns of h rimp , we have a po tenti a l fi shing a rea of a pproxi mate ly 250,000 km 2 . If yields o n the o rd er of 1-10 metri c tons/km2 are rea li st ic, then it seems probabl e that H . ensiJe ,. a nd o th e r members of thi s genus rep re ent a n unexp lo ited reso urce of conside ra bl e magnitude .

4 Da ta fr om Pascagoula F ish ery Produ cts T ec hnology Laboratory , Pascagoula, MS 39567. S L. G . Eld ri dge , University of G u am, Agana, Gu am 969 10. Pers. comm .

ACKNOWLEDGMENTS

We th a nk William Van Heukelem fo r prov iding catch data obtained with the FV Va/hllH Maid and Bernard Ito fo r th e photograph of the traps. Thomas A . Clarke and Lael L. Ron­ho lt made pertinent comment on the ma nuscript. Richard B. Roe provided ex plo ra to ry trawling da ta for the west­ern central Atlantic . The illustrations a re by Tamotsu Nakata.

L1TERATU RE CITED Butler, T . H . 1963. An improved prawn

trap. Fish . Res . Board Can. , BioI. Stn ., Nanaimo, C irc . 67, 7 p.

C larke, T . A. 1972 . Explo ration fo r deep benthic fish and crustacean re ources in H awai i. U ni v. Hawaii , H awai i Ins!. Mar. Bi oI. T ech . Rep . 29 : 1- 18.

Gulland , J . A. (edito r). 197 1. Th e fi sh re­sources of th e ocean. Fish. News (Books) . Ltd ., Surrey, 255 p.

H ancock, D. A., and G . Henriqu ez . 1968. Stock assessment in th e shrimp ( H el ero­ca rpus r eed i) fi shery of C hil e. In M. . Mistakidis (editor), Proceedin gs of th e World Scientific Conference on th e Biol­ogy and Cultur e of Shrimps and Pr awns held in Mexico C ity, Mexico, 12-21 Jun e 1967 . FAO Fish. Rep. 57, 2:443-465.

Kess ler , D. W. 1969. Test-tank studi es of shrimp -pot effi ciency . Fish . Ind . Res. 5: 15 1- 160.

Moiseev, P. A. 1969. Biologicheski e resursy Mirovogo okeana (Th e living resources of th e world ocean). Izd atel 's tvo "Pishche­vaya Promyshlennost" Moskva. (Trans­lated by I r ae l Program Sci. Transl. , 1971 , 334 p. ; ava ilabl e U .S. Dep . Comm er., Natl. T ec h. Inf. Serv., Springfi eld, V A, as TT 7 1-50026) .

Yoshida , H. O. 1972. Exploratory bo ttom trawling in H awaii an waters. Indo-Pac . Fish . Coun c ., Proc., 13 Sess., Sect. 3:255 -26 1.

MFR Paper 1095. From Marine Fisheries Review , Vol . 36, No . la,

Octo ber 1974 . Cop ies o f th is paper, in lim i ted numbers , are avail­ab le from 083 , Techni cal Information Divis ion , Envi ronmental SCience Informat io n Cen ter, NOAA , Wash ington , DC 20235.

30